EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED

STEEL TUBE COMPOSITE BEAMS

KHAWLA A. FARHAN1, MUHANED A. SHALLAL2

The main objective of this study is to highlight the performance of beams composed of lightweight concrete-
filled steel tubes (square and circle sections) composite with reinforced concrete deck slab. A total of nine
composite beams were tested included two circular and seven square concrete-filled steel tubes. Among the nine
composite beams, one beam, S20-0-2000, was prepared without a deck slab to act as a reference specimen. The
chief parameters investigated were the length of the specimen, the compressive strength of the concrete slab, and
the effect of the steel tube section type. All beams were tested using the three-point bending test with a
concentrated central point load and simple supports. The test results showed that the first crack in the concrete
deck slab was recorded at load levels ranging from 50.9% to 77.2% of the ultimate load for composite beams
with square steel tubes. The ultimate load increased with increasing the compressive strength of the concrete
slab. Shorter specimens were more stiffness than the other specimens but were less ductile. The slip values were
equal to zero until the loads reached their final stages, while the specimen S20-55-1100 (short specimen)
exhibited zero slip at all stages of the load. The ultimate load of the hollow steel tube composite beam was
13.2% lower than that of the reference beam. Moreover, the ductility and stiffness of the beam were also higher
for beams with composite-filled steel tubes.

Keywords: Concrete-filled steel tube (CFST), Lightweight concrete, Composite beam, Perfobond-rib shear
connector, Reinforced concrete, Compressive strength, Cracks

1
University of Al-Qadisiyah, Civil Engineering Department, Al-Diwania, Iraq, e-mail: ms.Khawla.farhan@qu.edu.iq
2
Prof., PhD., Eng., University of Al-Qadisiyah, Civil Engineering Department, Al-Diwania, Iraq, e-mail:
mohanad.shallal@qu.edu.iq
236 K.A. FARHAN, M.A. SHALLAL

1. INTRODUCTION

A composite beam makes use of the material properties of both steel and concrete. It is more
ductile, has higher strength, and is cheaper than its steel equivalent [1]–[3]. Concrete and steel
together as a composite in the construction enhances the ultimate capacity and stiffness. The
concrete filling in the steel tube not only prevents the local buckling of the steel but also raises the
ductility of the concrete-filled steel tube CFST to higher and better levels up to the ultimate load
[4], [5]. Recently, the use of CFST has become common in buildings, bridges, and other structures
[6], [7]. The use of CFST members in moment-resisting frames reduces the need for incorporating
stiff elements in panel zones and zones that demanded high strain [8]. Bridges with CFST members
are estimated to reduce noise and vibration levels compared to those with pure steel members [9].
Also, CFST members have been proven to be cost effective in building structures [10]–[12]. Some
usages of CFST members are shown in Fig. 1. Under bending loads, the CFST members perform
better than hollow tubes [13]. The lateral confinement by the steel tube adds to the strength of the
concrete core [14].
Most previous studies were focused on columns, as they are critical structural elements in high-rise
buildings and in areas prone to earthquakes. Other applications, such as flexural members (CFST)
are few in the research field. Han [15] did a study on the flexure behavior of CFST. Han tested 36
composite beam specimens that were filled with self-compacting concrete (SCC). Many parameters,
such as the ratio of shear span to depth (1.25-6), steel yielding strength (235-282 MPa), sectional
types (circular and square), and the ratio of tube diameter or width-to-wall thickness were used in
this study. Han found that the CFST filled with SCC are similar to the composite beams that are
filled with normal concrete. Moreover, the effect of the shear span to depth ratio was not evident on
the square and circular CFST.
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 237

Fig. 1. Applications of CFST composite members.

The flexural capacities of the section were predicted using the four design codes and the method
proposed by Han [15], [16]. These methods were compared with the test results by using the
proposed method of Han in his study, and the recommendation of design codes. Han’s method was
the best-predicted method. Shallal [17] tested twelve CFST beams in a study the main objective of
which was to determine the maximum load of CFST beams (square and circular) filled with
different types of concrete under flexural effect. The study presented an experimental test on the
flexural performance of square and circular CFST subjected to bending. From the results of the
investigation, it was observed that the maximum moment strength of the CFST was generally
higher than that of the corresponding hollow tube beams. This improvement in strength was
attributed to the steel confinement for the concrete core which increased the ductility and moment
capacity of the composite beam. Zhongqiu [18] tested a composite beam which consisted of a
lightweight CFST and a reinforced concrete deck slab. Bending tests were carried out on the beam
to study its performance. On the basis of the test results and the assumptions made in the study, the
suggested method of calculation was proved by finite element analysis. The results displayed that
the composite beam with slab deck and lightweight CFST had good ductility and high capacity. The
slip between the CFST and deck slab was ignored in this study. The flexural capacity for such a
composite beam by taking into account the yield strength of the steel tube and the concrete strength
can be evaluated by the force balance method. The superposition of the deck slab and lightweight
CFST caused a higher flexural stiffness in the elastic stage. Mossahebi et al. [19] carried out
238 K.A. FARHAN, M.A. SHALLAL

experimental studies to understand the performance of a bridge girder consisting of a steel tube
filled with concrete and made composite with a concrete deck. The test specimen displayed good
ductility and maintained its strength up to the end of the test. Results of this study demonstrated the
possibility of using a concrete-filled tube as a bridge girder. The practical application of the new
type of CFST girder is presented by Kang et al. [20]. The flexural behavior of the CFST girder that
was experimentally tested in this study, varied according to the strength of the filling material. Their
application as continuous girders was also evaluated. Results of the test showed that CFST girders
had good ductility and maintained their strength up to the end of the loading. Results of this
investigation demonstrated the favorable potential of the concrete-filled tube as a bridge girder.
Junghyun et al. [21] presented a study on CFST girders. This study included an experimental part
that included examining three specimens. The theoretical part included the analysis of the
specimens by the finite elements using the Abaqus program, and suggest new equations to find the
flexural capacity of a CFST composite girder.
This study presents the investigation of nine composite CFST beams with a deck slab. The
parametric study includes different lengths of specimens, section type of the steel tube, and three
types of normal concrete used in the concrete deck slab. Perfobond strip was used as a shear
connection key in the composite beam, and there are many applications of perfobond connectors
that are used to connect the CFST members with the concrete beam. The shape of the perfobond
connector that was used in the study is shown in Fig. 2. One mixture of lightweight aggregate
concrete was used to fill the hollow beam sections by using clay aggregate (LECA), and three
strengths of normal concrete were used in the concrete deck slabs. Not many studies have been
conducted on CFST as a flexural member. This study presents the application of this new sort of
CFST beam composite with a concrete deck slab. The concrete compressive strength used in filling
the steel tube had almost no effect on the flexural strength, therefore, lightweight concrete was used
to produce a relatively light composite beam in this study. Nine specimens were designed through
which the effect of change in the length of the specimen with fixed depth was evaluated for three
specimens. The effect of concrete compressive strength was used in the deck slab where three
different mixes were also studied. The effect of the different sections of the steel tube was
considered in the study where two specimens type were used, one with a circular section and the
other with a square section. The use of a deck slab was also studied where one of the specimens was
used without a deck slab and the effect of filling the tube with concrete as one of the specimens
formed was without concrete.
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 239

2. EXPERIMENTAL PROGRAM

Nine specimens were casted. Each specimen consisted of two main parts: the deck slab and the
CFST beam. Seven of the CFST beams were square sections and the other two were circular
sections. A summary of specimens is listed in Table 1. The nomenclature of the specimen contains
four parts. The first is S or C which refers to a section type (the S means square, and C is for
circular). The first number refers to the strength of lightweight concrete inside the tube (20 and 0);
zero means that there was no concrete inside the steel tube. The second number refers to the
strength of the concrete slab deck (55, 45, and 35 MPa). The last number refers to the total length
of specimens: 1100, 2000, and 2900 mm. The dimensions of the steel tube were 150×150×4 mm,
while the dimensions of the circular section were 150 mm in diameter and 3.8 mm in thickness.
Fig. 3 shows the details of these specimens. The deck slab dimensions were 450×75 mm which
included a steel reinforcement with a diameter of 8 mm. The detail of reinforcement is shown in
Fig. 3. The dimensions of the perfobond connector were 50×8 mm which was used as a shear
connector between the steel tube and the deck slab. The bottom layer of transverse steel
reinforcement passed through the perfobond plate that had a hole of 10 mm. One concrete mix is
used to fill the steel tubes (lightweight concrete 20 MPa), while just one square specimen has no
infilled concrete (hollow steel tube). The strengths of the concrete deck slab of specimens were (55,
45, and 35 MPa).

Fig. 2. Composite beam with perfobond connector.

240 K.A. FARHAN, M.A. SHALLAL

Table-1 Specimen properties.

Specimen label Effective length Section type #
S20-55-2000 1800 mm square
S20-55-1100 900 mm square
S20-55-2900 2700 mm square
S0-55-2000 1800 mm square
S20-45-2000 1800 mm square
S20-35-2000 1800 mm square
S20-0-2000 1800 mm square
C20-55-2000 1800 mm circular
C20-35-2000 1800 mm circular
# square (150×150×4 mm), circular (Φ=150, t=3.8 mm)

3. MATERIAL PROPERTIES

All concrete mixtures contained ordinary Portland cement (Type I), with 10 mm as the maximum
size of the coarse aggregate, while the fine aggregate conformed to the requirements of ASTM-
C330 specification. The concrete filled in tubes was a lightweight concrete that used gravel with a
maximum size of 10 mm as clay aggregate (LECA). Details of concrete mixtures and cubic
compressive strengths are summarized in Table 2. The values of yield stress and ultimate tensile
strength of the steel tubes, steel reinforcement, and perfobond connector are summarized in Table 3.

4. TEST ARRANGEMENT

All specimens were tested using a universal testing machine. The composite beams were tested
under concentrated loads in the middle of the span. Fig. 4 shows a diagram of the device, load cell,
and locations of LVDT sensors.
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 241

Concrete deck slab

Concrete deck slab

Steel tube

Light weight concrete

Light weight concrete
Steel tube

A- 3D composite beam

Deck slab Deck slab

Perforated shear connector (50x Perforated shear connector (50x
8 mm) 8 mm)

450 450
75

75
) 8 @ 150 ) 8 @ 150
4 ) 8 mm 4 ) 8 mm
150

150
top and bottom top and bottom
Lightweight Lightweight
concrete concrete

Steel pipe (t=3.8 mm) Steel tube (150x150x4)

B- Beam cross section

Deck slab

Concrete filled steel tube

Support
100 L-200 100

C- Side view of concrete filed steel tube with deck slab

Fig. 3. Scheme of the specimen.

242 K.A. FARHAN, M.A. SHALLAL

Table-2 Mixing concrete properties.

Superplasticizer Compressive strength

Mix Mixing ratio W/C ratio
(SP.) % (MPa)
Lightweight concrete (20 MPa) 1:1.75:0.5 0.3 1 21.85
Normal concrete, M2: 55 MPa 1:1.47:2.2 0.3 0.5 52.17
Normal concrete, M3: 45 MPa 1:1.89:2.84 0.4 0.3 47.93
Normal concrete, M4: 35 MPa 1:2.67:4 0.5 - 35.14

Table 3 Steel members properties.

Yield stress, Fy Ultimate strength, Fu

Component
(MPa) (MPa)
Square tube 390 433
Circular tube 305 391
Reinforcement 464 606
Perfobond connector 357 370

Hydraulic
jack

Load cell

LVDT3 Composite beam

support
support
LVDT2 LVDT1

(A)

Load cell

C20-35-2000
LVDT3

(C)
LVDT2
LVDT1

(B)
Fig. 4. Scheme of universal test machine and measurement settings.
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 243

5. RESULTS AND DISCUSSION

The results obtained from the tests showed that all the specimens failed in the bending test except
the specimen S20-55-1100 which exhibited a different behavior in failure. The results are listed in
Table 4. Fig. 5 shows some pictures of the test specimen. Ultimate load, cracks, load-deflection
curve, and slip at the end of the composite beam were also determined from the tests.

5.1 CRACKS
Cracks inside the steel tube could not be seen during the test. The only cracks that could be seen
were those that were in the concrete deck slab. Table 4 lists the loads (Pcrack) that were recorded
when the first crack appeared in the concrete deck slab. The appearance of the first crack in the
concrete deck slab was recorded between 50.9% and 77.2% of the ultimate load for the specimens
having a square steel tube. It can be concluded that the neutral axis existed in the zone of the steel
tube before the load was applied. As a result of cracks in the concrete inside the steel tube, the
neutral axis moved upwards within the concrete deck slab. The first crack was noticed in the
concrete deck slab of the specimen S0-55-2000 at 106.2 kN, which represents 50.9% of the ultimate
load. This percentage was the lowest among all specimens owing to the lack of concrete inside the
steel tube. The lack of concrete inside the steel tube made it less stiff compared to the other
specimens, so the neutral axis moved to the concrete deck slab in the early stage of the load. Three
specimens, S20-35-2000, S20-45-2000, and S20-55-2000, were studied for determining the effect
of change in the compressive strength of the deck slab. The first crack was observed in the concrete
deck slab of these specimens at loads 126.04, 160.01, and 172.25 kN, respectively, which represents
55.5, 69.2, and 71.7%, respectively, of the ultimate loads of these specimens. It was observed that
the specimen which had a high compressive strength imparted a higher stiffness in the specimens
and, hence, delayed the crack formation in the concrete deck slab. Three specimens, S20-55-2900,
S20-55-2000, and S20-55-1100, were tested for determining the effect caused by the specimen’s
length. The first crack was observed in the concrete deck slab of these specimens at loads of 105.3,
172.25, and 400.27 kN, respectively, which represent 65.9, 71.7, and 77.2%, respectively, of the
ultimate loads of these specimens. These results showed that there existed an inverse relationship
between the percentage of the first crack load to the ultimate load and the length of the specimen.
The increase in specimen length led to a decrease in the percentage of the crack load. To study the
concrete cracks inside the steel tube, a section of the steel tube was cut and removed so that
observations of the cracks in the concrete part which was used to fill the steel tube could be
244 K.A. FARHAN, M.A. SHALLAL

recorded. Fig. 6 shows the cracks in these specimens. At the corners of the square tube, part of the
concrete remained adhered to the tube, as shown in Fig. 6. It can be seen that the cracks in the
circular section were more in number and had less width than the cracks in the square section. The
cracks of the concrete deck slab were fewer than the cracks in the concrete inside the tube. This is
normal because higher stress is present near the bottom section. As for the width of the cracks, the
cracks of the concrete deck slab were found to be more extensive than the cracks in the concrete
inside the steel tube. This explains the presence of a larger number of cracks in this zone as the
stress is moderately distributed because of concrete surrounding the steel section. The specimen,
S20-55-1100, showed two types of cracks in the concrete deck slab; first, there were flexural cracks,
but at the failure point, a crack resulting from the shear stress also started becoming apparent, as
shown in Fig. 6.

Table 4. Result of the tested composite beams.

Stiffness
Ultimate load Pcrack Pcrack/ Pu Pslip Pslip/Pu
Beam (kN/m
(Pu) (kN) (kN) (%) (kN) (%)
×103)
S20-55-2000 240.25 172.25 71.7 215.39 89.7 52.71
S20-55-1100 518.59 400.27 77.2 - - 74.11
S20-55-2900 159.77 105.30 65.9 112.84 70.6 11.02
S0-55-2000 208.57 106.21 50.9 191.05 91.6 36.87
S20-45-2000 231.31 160.01 69.2 215.04 93.0 44.41
S20-35-2000 227.11 126.04 55.5 116.65 51.4 29.85
S20-0-2000 117.92 - - - - 12.97
C20-55-2000 208.18 159.96 76.8 193.56 93 44.63
C20-35-2000 193.59 116.03 59.9 175.73 90.8 29.75

Pcrack: load at first crack in the concrete deck.

Pslip: load at first slip between the concrete deck and steel tube at the end of the specimen.

5.2 ULTIMATE LOAD AND LOAD-DEFLECTION CURVES

5.2.1 EFFECT OF COMPRESSIVE STRENGTH

Table 4 shows the ultimate load values for all specimens. The results showed that the ultimate load
was reduced by decreasing the compressive strength of concrete. Also, the ultimate loads of
specimens S20-45-2000, S20-35-2000, and S20-0-2000, were less than the ultimate load for
specimen S20-55-2000 by 3.7, 5.5, and 50.9%, respectively. The last percentage, 50.9%, for
specimen S20-0-2000 showed a large decrease because of the absence of a concrete deck slab.
Therefore, the stiffness was lower than the other specimens. Fig. 7 shows the load-deflection curves
owing to the change in the concrete compressive strength in the concrete deck slab. From this
figure, a slight difference in the stiffness of the specimens can be noticed except in the case of
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 245

specimen S20-0-2000 which showed a significant difference in stiffness because of the absence of a
concrete deck slab. Two specimens, C20-55-2000 and C20-35-2000, illustrated the effect of change
in the compressive strength of concrete in the deck slab. The ultimate loads of these specimens
confirmed that the ultimate load was affected by concrete compression strength as the ultimate load
decreased by decreasing the compression strength of concrete. Fig. 8 shows the load-deflection
curves. It is evident from the figure that the specimens exhibit the same behavior under the load.

5.2.2 EFFECT OF LENGTH

The results obtained from the tests showed that the beams having different lengths (S20-55-1100,
S20-55-2000, and S20-55-2900) with the same properties exhibit different behavior as shown in
Fig. 9. The ultimate loads for specimens S20-55-1100, S20-55-2000, and S20-55-2900 were
518.59, 240.25, and 159.77 kN, respectively. These results were as expected because of the
difference in lengths of the specimens. The results based on moments were compared, as shown in
Fig. 10. The maximum moment for specimens S20-55-1100, S20-55-2000, and S20-55-2900 were
116.68, 108.11, and 107.84 kN, respectively. As shown in Figs. 9 and 10, the beam, S20-55-1100,
had more stiffness but less ductility than the two other beams, S20-55-2000, and S20-55-2900.
Thus, the beam S20-55-1100 failed by shear and flexure, as shown in Fig. 6, while beams S20-55-
2000, and S20-55-2900 failed in flexure.

C20-35-2000
S20-0-2000

S20-55-1100

S20-35-2000

S20-55-2900

Fig. 5. Specimens during and after the test.

246 K.A. FARHAN, M.A. SHALLAL

Fig. 6. Cracks in the test specimens.

Fig. 7 Effect of concrete compressive strength in Fig. 8 Effect of concrete compressive strength in
specimens of the square tube on load-deflection specimens of the circular tube on load-deflection
curves. curves.
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 247

Fig. 9. Effect of specimen length on load-deflection Fig. 10. Effect of specimen length on moment-
curves. deflection curves.

5.2.3 EFFECT OF FILLING CONCRETE

In this section, a specimen S20-55-2000 with concrete strength of 55 MPa for a deck slab, and
lightweight concrete filled in the tube (21.17 MPa), and another specimen, S0-55-2000 with
concrete strength of 55 MPa for a deck slab but with a hollow steel tube was considered. In the mid-
span of the beam S0-55-2000, the bottom flange buckled upward as shown in Fig. 11. The ultimate
load for the CFST beam S20-55-2000 was 240.25 kN, while for the CFST beam S0-55-2000 it was
208.57 kN, a decrease of 13.2% compared with the reference beam S20-55-2000. Fig. 12 shows the
load-deflection curves. It can be seen from the curve that concrete filling the tube significantly
increased the ductility. Consequently, the stiffness also increased.

5.2.4 EFFECT OF THE TUBE SECTION TYPE

In this study, a comparison among the CFST beams having different cross-sections of steel tubes
(square or circle) has been made. The specimen S20-35-2000’s ultimate load, 227.11 kN, was
14.8% higher than the ultimate load of 193.59 kN for specimen C20-35-2000. The specimen S20-
55-2000’s ultimate load, 240.25 kN, was 13.4% higher than the ultimate load of 208.18 kN for the
specimen C20-55-2000. This is shown in Fig. 13.

Fig. 11. Buckling of the bottom flange for specimen S0-55-2000.

248 K.A. FARHAN, M.A. SHALLAL

Fig. 12. Effect of filling concrete on load- Fig. 13. Effect of tube section type on load-
deflection curves. deflection curves.

5.3 SLIP

In this section, the slip between the concrete deck slab and steel tube was measured using LVDT3 at
the end of the specimen, as shown in Fig. 4. Slip can be defined as a difference in horizontal
movement between the CFST and the concrete deck slab in their interface zone [22], [23]. As
shown in Table 4, the slip for all beams began at 90% of ultimate load except two beams, namely,
S20-55-2900 which began at 70.6% of ultimate load and S20-35-2000 which began at 51.4% of the
ultimate load. Figs. 14-17 show a similar behavior of slip at the end of the beam during increased
loads but with slight differences. Figs. 14-17 show that there was no slipping at the end of the beam,
and the slip value was zero until the loads reached their final stages. Table 4 shows that the
percentage of load in which a slip appeared, ranges from 89.7 to 93.0%, except for three specimens:
S20-55-1100, S20-55-2900, and S20-35-2000.
Fig. 14 shows that the specimen S20-35-2000 behaved differently from other specimens as the slip
started at an early stage. The slip between concrete and CFST started at 51.4% of the ultimate load
and then increased with increasing load until the slip values reached values comparable with the rest
of the specimens. This might be due to the decrease in the compressive strength of this specimen.
However, specimen C20-35-2000 showed no such behavior, as shown in Fig. 15. Fig. 16 shows that
the specimen S20-55-1100 had a higher ultimate load and zero slip. The reason can be attributed to
the short length of the specimen and, therefore, curvature was found to be insignificant. Specimen
S20-55-2900 showed a behavior similar to the rest of the specimens, but the slip value began at
70.6% of the ultimate load. This might be caused by the length of the specimen compared with the
rest of the specimens.
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 249

Fig. 14 Effect of concrete compressive strength

Fig. 15 Effect of section type on load-slip curves.
on load-slip curves.

Fig. 16 Effect of specimen length on load-slip Fig. 17. Effect of filling concrete on load-slip
curves. curves.

6 CONCLUSION
The following comments can be drawn as the most important concluding remarks from the
experimental work conducted in this research as reported above.
1. The appearance of the first crack in the concrete deck slab was recorded between 50.9% and
77.2% of the ultimate load for the composite beams with a square steel tube.
2. The close examination of the beams after testing showed that the cracks of the lightweight
concrete filling in circular steel tubes were higher in number but smaller in width than the cracks
in square tubes.
3. The ultimate load capacity of the tested beams was reduced by 3.7% when the compressive
strength of the deck slab was decreased from 55 to 45 MPa, while a higher decrease of 5.5% was
recorded when the compressive strength was further reduced to 35 MPa.
250 K.A. FARHAN, M.A. SHALLAL

4. The short beam, S20-55-1100, showed higher initial stiffness than the other beams S20-55-2000,
and S20-55-2900, by 28.9% and 85.1%, respectively. However, its deflection corresponding to
the ultimate load was lower by 105.0% and 135.1%, respectively.
5. The composite beams with a square steel tube exhibited higher ultimate load and stiffness than
those of beams with a circular steel tube.
6. Excluding the three beams S20-55-1100, S20-55-2900, and S20-35-2000, the slip values were
zero until the loads reached their final stages. The load at which the slip appeared for all other
beams ranged from 89.7 to 93.0% of the ultimate loads. For the beams S20-55-2900 and S20-35-
2000, the slip was recorded at 70.6% and 51.4% of the ultimate loads, respectively, while zero
slip was recorded for the beam S20-55-1100.
7. The ultimate load capacity for the empty steel tube beam decreased by 13.2% compared with the
reference beam. Similarly, the initial stiffness of this beam was lower than that of the reference
beam by 30.1%.

REFERENCES
[1] B.-C. Chen and T.-L. Wang, “Overview of concrete filled steel tube arch bridges in china,” Pract. Period.
Struct. Des. Constr., vol. 14, no. 2, pp. 70–80, 2009.
[2] Y. Liu, L. Guo, and Z. Li, “Flexural behavior of steel-concrete composite beams with U-shaped steel girders,”
in Proceedings 12th international conference on Advances in Steel-Concrete Composite Structures - ASCCS
2018, 2018.
[3] J. Derysz, P. M. Lewiński, and P. P. Więch, “New concept of composite steel-reinforced concrete floor slab in
the light of computational model and experimental research,” Procedia Eng., vol. 193, pp. 168–175, 2017.
[4] L.-H. Han, F.-Y. Liao, Z. Tao, and Z. Hong, “Performance of concrete filled steel tube reinforced concrete
columns subjected to cyclic bending,” J. Constr. Steel Res., vol. 65, no. 8–9, pp. 1607–1616, 2009.
[5] A. W. Al Zand, W. H. W. Badaruzzaman, A. A. Mutalib, and S. J. Hilo, “Flexural behavior of CFST beams
partially strengthened with unidirectional CFRP sheets: experimental and theoretical study,” J. Compos.
Constr., vol. 22, no. 4, pp. 1-16, 2018.
[6] S. Fujikura and M. Bruneau, “Dynamic analysis of multihazard-resistant bridge piers having concrete-filled
steel tube under blast loading,” J. Bridg. Eng., vol. 17, no. 2, pp. 249–258, 2012.
[7] A. K. Kvedaras, G. Šaučiuvėnas, A. Komka, and A. Juozapaitis, “Design of circular composite beams with a
different concrete core considering the effect of concrete in tension,” J. Civ. Eng. Manag., vol. 22, no. 1, pp.
118–123, 2015.
[8] M. Elchalakani, X. L. Zhao, and R. H. Grzebieta, “Concrete-filled circular steel tubes subjected to pure
bending,” J. Constr. Steel Res., vol. 57, no. 11, pp. 1141–1168, 2001.
[9] S. Nakamura, Y. Momiyama, T. Hosaka, and K. Homma, “New technologies of steel/concrete composite
bridges,” J. Constr. Steel Res., vol. 58, no. 1, pp. 99–130, 2002.
[10] J. Webb and J. J. Peyton, “Composite concrete filled steel tube columns,” Second Natl. Struct. Eng. Conf. 1990
Prepr. Pap., p. 181, 1990.
[11] J. Flor, R. Fakury, R. Caldas, F. Rodrigues, and A. Araújo, “Experimental study on the flexural behavior of
large-scale rectangular concrete-filled steel tubular beams,” Rev. IBRACON Estruturas e Mater., vol. 10, pp.
895–905, 2017.
[12] “Performance of concrete filled steel tube (CFST) section: a review,” Int. J. Sci. Res., vol. 4, no. 11, pp. 645–
647, 2015.
[13] L.-H. Han, W. Li, and R. Bjorhovde, “Developments and advanced applications of concrete-filled steel tubular
(CFST) structures: Members,” J. Constr. Steel Res., vol. 100, pp. 211–228, 2014.
[14] Y.-F. An, L.-H. Han, and C. Roeder, “Flexural performance of concrete-encased concrete-filled steel tubes,”
Mag. Concr. Res., vol. 66, no. 5, pp. 249–267, 2014.
 EXPERIMENTAL BEHAVIOUR OF CONCRETE-FILLED STEEL TUBE COMPOSITE BEAMS 251

[15] L.-H. Han, H. Lu, G.-H. Yao, and F.-Y. Liao, “Further study on the flexural behaviour of concrete-filled steel
tubes,” J. Constr. Steel Res., vol. 62, no. 6, pp. 554–565, 2006.
[16] L.-H. Han, “Flexural behaviour of concrete-filled steel tubes,” J. Constr. Steel Res., vol. 60, no. 2, pp. 313–337,
2004.
[17] M. A. Shallal, “Flexural behavior of concrete-filled steel tubular beam,” in 2018 International Conference on
Advance of Sustainable Engineering and its Application (ICASEA), 2018, pp. 153–158.
[18] Z. Fu, Q. Wang, Y. Wang, and B. Ji, “Bending performance of lightweight aggregate concrete-filled steel tube
composite beam,” KSCE J. Civ. Eng., vol. 22, no. 10, pp. 3894–3902, 2018.
[19] N. Mossahebi, A. Yakel, and A. Azizinamini, “Experimental investigation of a bridge girder made of steel tube
filled with concrete,” J. Constr. Steel Res., vol. 61, no. 3, pp. 371–386, 2005.
[20] J. Y. Kang, E. S. Choi, W. Chin, and J. W. Lee, “Flexural behavior of concrete-filled steel tube members and its
application,” Int. J. Steel Struct., vol. 7, pp. 319–324, 2007.
[21] J. Cho, J. Moon, H.-J. Ko, and H.-E. Lee, “Flexural strength evaluation of concrete-filled steel tube (CFST)
composite girder,” J. Constr. Steel Res., vol. 151, pp. 12–24, 2018.
[22] M. A. Shallal, A. M. K. Al Musawi, and F. I. Muss, “Non-linear analysis of continuous composite beam
subjected to fire,” International Journal of Civil Engineering and Technology (IJCIET), vol. 9, pp. 521-532.
[23] M. Shallal and A. M. K. Al Musawi, “Non-linear analysis of composite beam subjected to fire,” J. of
Engineering and Applied Sciences, vol. 13, pp. 9643–9650.

LIST OF FIGURES AND TABLES:

Fig. 1. Applications of CFST composite members.

Fig. 2. Composite beam with perfobond connector.
Fig. 3. Scheme of the specimen.
Fig. 4. Scheme of universal test machine and measurement settings.
Fig. 5. Specimens during and after the test.
Fig. 6. Cracks of the testing specimens.
Fig. 7 Effect of concrete compressive strength in specimens of the square tube on load-deflection curves.
Fig. 8 Effect of concrete compressive strength in specimens of the circular tube on load-deflection curves.
Fig. 9. Effect of specimen length on load-deflection curves.
Fig. 10. Effect of specimen length on moment-deflection curves.
Fig. 11. Buckling of the bottom flange for specimen S0-55-2000.
Fig. 12. Effect of filling concrete on load-deflection curves.
Fig. 13. Effect of tube section type on load-deflection curves.
Fig. 14 Effect of concrete compressive strength on load-slip curves.
Fig. 15 Effect of section type on load-slip curves
Fig. 16 Effect of specimen length on load-slip curves.
Fig. 17. Effect of filling concrete on load-slip curves.
Table-1 Specimen properties.
Table-2 Mixing concrete properties.
Table 3 Steel members properties.
Table 4. Result of the tested composite beams.

Received 11.10.2019 , Revised 22.02.2020